Kidney International, Vol. 13 (1978), pp. 15-26

Pharmacological mechanisms of analgesic nephropathyJULIAN H. SHELLEY

C. H. Boehringer Sohn, Bracknell, Berkshire, United Kingdom

The pre-Socratic philosophers were reputed to beable to entertain two mutually contradictory beliefswith apparent equanimity. In reading this review,which undoubtedly reflects the bias of the writer and,in turn, the authors of the cited works, the complexi-ty and contradictory nature of much of the data willbe evident; one cannot view, however, the resultwith any degree of equanimity. Much remains to bedone in the experimental field before a unifying con-cept on the pathogenesis, incidence, and implica-tions of analgesic nephropathy can be enunciated.

Scope of the problem. The review of Gsell [11"From phenacetin nephritis to analgesic nephropa-thy" admirably reflects the change in attitude to theproblem, particularly in the last ten years. Sinceother investigators have indicated the current scopeof the problem [2—22], no further comment is need-ed. Adverse reactions referable to the kidney areattributed to aspirin (used alone and in combinationwith other agents), phenacetin (in combination), par-acetamol (alone and in combination), phenylbuta-zone, oxyphenbutazone, ketophenbutazone, sul-phinpyrazone, indomethacin, ibuprofen, ketoprofen,fenoprofen, tolmetin, glaphenine, niflumic acid,mefanamic acid, tolfenamic acid, bucloxic acid, ado-fenac, and naproxen. The data used in this revieware derived both from published reports and drug-monitoring committees (Table 1) [23—66]. In sum-mary, salicylates and phenacetin used in combinationand aspirin used alone are frequently implicated, thepyrazoles less so; and the newer nonsteroidal antiin-flammatory agents, e.g., the phenylalkanoic acidsand fenamic acids, play a small but significant part.Occasional reports implicating dextropropoxyphene,caffeine, and older analgesics are also noted.

Pharmacological and toxicological studies in animals andman

Short-term therapy. Indomethacin administration(150 mg/day, for three days) was shown to inducewater and sodium retention both in normal subjectsand in patients with kidney damage [73—75]. In

15

patients with preexisting reduction in glomerular fil-tration rate (GFR) and renal plasma flow (RPF), thisreduction was intensified, and sodium restriction ledto a more pronounced effect; normal subjectsshowed no change in GFR or RPF. In another study,indomethacin administration (150 mg/day) decreasedthe GFR by 35% and the RPF by 23% in nephroticpatients with impaired renal function.

Both phenylbutazone and ketophenbutazoneadministrations (single dose of 600 mg, either orallyor i.m.) were shown to impair '311-hippuran excre-tion, which was interpreted as an acute impairmentof tubular function [76]. In another study [77], theisotope renogram was used to measure the changesinduced by oxyphenbutazone administration (750 mgdaily). These changes were interpreted as a reflectionof impaired blood flow, tubular function, and there-fore, clearance of the isotope.

In normal subjects, the short-term administrationof aspirin (oral doses of 5.5 g daily) was shown toproduce sodium retention, a decreased urine vol-ume, and tubular reasorption of free water; nochange in creatinine clearance was noted. The effectwas more pronounced when plasma renin had beenelevated by low sodium balance [73, 74]. In patientswith chronic renal insufficiency, aspirin i.v. (lysineconjugate, 750 mg) caused a decrease in sodium renalexcretion within 15 mm which lasted for six hours, adecrease in GFR to 54%, and a decrease in RPF to66% of control values [73].

Aspirin was shown to reduce the glomerular filtra-tion rate in 13 patients with renal impairment by 10%when given 20 mg/kg by mouth, while in a furthereight subjects [78], with no renal impairment, thelysine conjugate of aspirin (0.9 g, i.v.) produced asignificant reduction in GFR and PAH clearance inseven [79]. The effect was transient, however. Theblood salicylate concentration at the time of maximaleffect was 12.4 mg/l00 ml. The effect of aspirin in

0085-2538/78/0013-0015 $02.40© 1978, by the International Society of Nephrology.

16 Shelley

Table 1. Data for agents which have been correlated with adverse reactions reibrable to the kidneya

Phen- Para- Phenyl- Oxyphen- Indo- Keto- Mefanamic FlufenamicAspirinc acetin" cetamol' butazone' butazone methacin5 Ibuprofenh profen' acid' acid Aclofenac5 Naproxen

Abnormal urine 1 2 2Acute renal failure 3(2) 5(4) 2 2Abnormal function 1 2 1Abnormal and

deteriorating function 3(2) 2(1) 1(1) 1(1)Renal papillary necrosis 27(16) 52(40) 6(4) 4(2) 1(2) 3(2) 2(1)Tubular necrosis 2(1) 2(1)Hematuria 4 7 5 6 I 2Interstitial nephritis 1

Glomerular nephritis 1Albuminuria 1 1 1 1

Toxic nephropathy 5(1)Nephrosis 2 1

Dysuria/oliguria Ill

Data provided by the Committee on Safety of Medicines [23] for period March, 1964, to December1975, except as follows: ibuprofen, 12/65—12/75;ketoprofen, 2/70—12/75; flufenamic acid, 6/67—12/75; aclofenac, 4/70—12/75; naproxen, 7/70—12/75.

Figures in brackets denote fatal outcome.For data on fenoprofen, see [44]; tolmetin, see [45]; glaphenine, see [50-60]; niflumic acid, see [61-63]; bucloxic acid, see [64, 65]See also [4, 24, 25].See also [26-29].See also [30].See also [22, 31-35].See also [36-38].See also [39, 40].See also [41-43].See also [66].See also [46, 47].

'See also [38, 48, 49].Bucloxic acid: See also [64, 65].

reducing the urinary excretion of xylose is probablydue to its reduction of GFR [801. Ingestion of aspirincauses a transient shedding of renal tubular cells intothe urine; excretion of enzymes, however, e.g., lac-tate dehydrogenase (LDH) and N-acetyl-/3-D-glu-cosaminidase (NAG), persists. NAG excretion is asensitive test of renal damage and correlates wellwith impairment of concentrating power and experi-mental papillary necrosis in dogs [81, 82]. Leath-wood and Plummer [83] gave 3 g of aspirin acutely tonormal subjects, and Burry et al [81, 821 adminis-tered 4 g of aspirin to both normal subjects andpatients with rheumatoid arthritis. Leathwood andPlummer produced persuasive evidence that the kid-ney is the main source of LDH and that both celluriaand rise of LDH levels indicate cell damage. Burryfound that NAG levels were elevated ten days afterdosing, all values being higher in patients than inhealthy subjects, indicating persisting tubular dam-age which, however, is considered to reflect a mini-mal impairment of tubular function.

Long term therapy. Retrospective studies of renalfunction in patients who have received salicylate forlong periods have been contradictory. Burry et al[81, 821 suggested that when a large amount of sali-cylate is consumed over many years, an appreciabledegree of tubular function impairment results.Dieppe et a! [84] found, however, that in 8 of 20patients with rheumatoid arthritis who had receivedno drug treatment, the NAG excretion was abnor-

mally high, so that rheumatoid arthritis per se maybe important. Salant [85] found that in patients withanalgesic nephropathy (drugs not stated), distal tubu-lar function was impaired out of all proportion to thedegree of reduction in the glomerular filtration rate.The conflicting and inconclusive data from renallunction studies in rheumatoid arthritis has beenreviewed by Nanra and Kincaid-Smith [9].

The New Zealand Rheumatoid Association study[86] exonerated aspirin as a cause of nephrotoxicityon the basis of "a renal score" and creatinine clear-ance tests, but no measurement of tubular functionwas made. In contrast to the findings of Edwards eta! [871. Burry [81] detected no relation betweenglomerular function (creatinine clearance) and thedose of aspirin consumed; endogenous creatinineclearance, however, although inappropriate as a testof renal tubular function, should not be interpretedas an index of GFR when salicylates are adminis-tered [881. The salicylate itself probably disturbs theprotein binding of creatinine or affects its tubularsecretion. The renal clearance of Cr-edetic acidwas normal in subjects who showed a reduction increatinine clearance; so the test is not suitable inassessing renal function in patients receivingsalicylate.

In sharp contrast to the findings of Berg [73, 741,Elliot and Murdaugh [891, and others, patients withanalgesic nephropathy (compound analgesics), i.e.,where drugs have been administered chronically,

Analgesic nephropathy. Drug mechanisms 17

show significant sodium loss (i.e., the urinary sodiumis inappropriately high). Such depletion will, in turn,impair renal function, and this would correlate withthe known loss of long nephrons, as the remainingshort nephrons have a lower capacity for sodiumabsorption [901.

Patients with analgesic nephropathy may sufferfrom diminished concentrating ability, and this ismost clearly shown when compound mixtures havebeen taken for long periods; the evidence implicatingaspirin has been briefly summarized above. Muchless data is available with regard to paracetamol;but in a thorough study, Edwards et al [911 foundno evidence of impaired function (including concen-trating power) in 18 patients with rheumatoid arthritiswho had consumed between 2 to 30kg of paracetamolin the past and at least one gram of paracetamol dailyper year prior to the examination.

Aspirin is under investigation in the secondaryprevention of myocardial infarction, and the mostdetailed study published to date is that of the Coro-nary Drug Project [921. Seven hundred and twenty-seven patients received one gram of aspirin daily for10 to 28 months. "Blood in urine" was reported in1.2%, compared to 0.3% of 744 patients receivingplacebo. This was not, however, confirmed by dip-stick measurement of hematuria (1.53% of 523patients receiving aspirin, compared with 2.08% of528 patients receiving placebo). Urine protein, how-ever, was higher in the aspirin group.

In a prospective study, Dubach, Rosner, and MUll-er [27] showed that concentrating power was signifi-cantly reduced and serum creatinine increased inhigh analgesic users. Bengston [93], however, con-sidered that the incidence of lesions in the study wasunderestimated, and it is difficult to reconcile theconclusions of the paper with the findings therein.

Pharmacological and toxicological studies in animals

Any pharmacological analysis must take accountof the relevance of the species employed, the dosesused, whether a dose-response curve is demonstra-ble, and finally, the experimental conditions underwhich the studies were conducted and how theresults have been interpreted.

Doses employed and relevant pharmacokinetics.Many studies have employed doses far outside therealm of pharmacology; In many cases we are deal-ing with toxicologic studies in the proper sense of theterm. With the wealth of data available, no clear-cutpicture emerges, apart from the fact that very highdoses of agents are required before any lesion isobserved in a consistent fashion.

By comparing the whole-body total salicylate con-centration in man and Charles Rivers rats, Philips,

Leeling, and Kowalski [441 calculated that singledoses of 24 and 60 mg/kg of body wt in the ratproduced whole-body total salicylate concentrationsthat were equivalent to the ingestion of 8 or 20 tablets(325 mg) of plain aspirin in man. No distinctionbetween parent and metabolite was made, however,and the general status of the rats was not reported.At this dose no lesions were seen, which parallels thefindings of Bokelman et al [94] in rats (with bufferedaspirin) and Mclver and Hobbs [95] in pigs given 1 g/kg daily for ten months. Nanra [71, 72], however,with intermediate dose levels of aspirin (200 mg/kg!day) given from 10 to 60 weeks under conditions ofdehydration, produced renal papillary necrosis in 7of 13 rats, and 5 showed changes typical of chronicinterstitial nephritis. No control group with freeaccess to fluid was included, however. High doses ofaspirin or APC have produced renal papillarynecrosis consistently (Nanra, Kincaid-Smith [70, 96,97]: 500 mg/kg aspirin, 900 mg/kg APC; Saker andKincaid-Smith [134]: 500 mg/kg/day phenacetin, 500mg/kg/day APC). The latter dose of APC is equiva-lent on a wt/wt basis to that taken by some patients,but it is excessive on the basis of calculations ofPhilips et al [44]. It should be noted enpassant thatthe general health of rats given aspirin or APC at thisdose level was "dreadful" (Nanra [71]).

In other studies, large doses of indomethacin (75mg/kg/day) and phenylbutazone (400 mg/kg/day) giv-en to female Wistar rats produced renal papillarynecrosis, while smaller doses of indomethacin (4 mg/kg) produced changes in the basement membrane ofthe glomerulus [98]. Phenylbutazone, when giventherapeutically to a dog at a dosage of 25 mg/kg,induced tubular damage [99]; 40 mg/kg given for 21days in the cat produced tubular necrosis. Enormousdoses (3,000 mg/kg/day) of phenacetin [100] wererequired to produce even modest tubular damage,while 3,000 mg/kg/day of paracetamol (Nanra andKincaid-Smith [96]) produced lesions in three of sev-en dehydrated rats. Dextropropoxyphene (100 mg/kg/day) produced renal papillary necrosis in one ofnine rats, but six showed medullary calcification[96].

Brown and Hardy [101] gave up to 1,300 mg/kgphenacetin to female Sprague Dawley rats for 51days and a similar dose of phenazone and up to 1,200mg/kg amidopyrine for 38 days. Phenazone causedpersistent celluria with some cellular infiltration;amidopyrine produced severe papillary necrosis butno celluria; phenacetin caused neither effect. Alldrugs impaired concentrating power in this experi-ment.

The effects of high doses of aspirin or APC in rats(Sprague-Dawley and Wistar) are enhanced by dehy-

18 Shelly

dration and oligemic shock [8, 70, 72, 97]. Arnold etaT [102, 103], however, gave single doses of 100 mgIkg to Wistar rats and observed transient corticaldamage; tubular regeneration commenced within 24hours and was complete in three days, despite con-tinuation of dosing where this was done. Whenlesions were observed at 12 hr, a dose response wasseen, notably in females (100 to 600 mg/kg/day aspi-rin). Some degree of protection from the single largedose of aspirin was seen up to three weeks afteradministration, and when 10 mg/kg was given dailyfor nine days, some protection against a repeat doseof 40 mg/kg was afforded. Arnold et al [103] report-ed, too, that pretreatment with phenobarbitoneappeared to reduce the severity of the lesions, but nodetails were given. Papillary necrosis was never seenin these studies, and alkalization had no effect on thecortical lesions. Lesions exclusively in the cortex ofthe kidney have been observed before in patients onsalicylate therapy [103].

Ibuprofen, at doses approaching the LD50, pro-duced dilation of the distal tubules [104]. In longterm toxicity studies in rats initially given 180 mg/kg,but reduced to 60 mg/kg per day because of theintestinal ulceration, no histologic damage wasdetected, but data available on medullary function isscanty [104].

Data on concentrations of aspirin in total bodydistribution, plasma, and renal tissue are conflicting:Leeling, Philipps, and Kowalski [105] found a linearrelation between aspirin dose, average renal tissueconcentration in mg/g of tissue, and plasma concen-tration in mg/ml when using doses of aspirin of 100mg/kg to 700 mg/kg (this last dose approached theLD50; 3 of 21 rats died). Much lower plasma andtissue concentrations were obtained when bufferedeffervescent aspirin was used—this preparation pro-duces an alkaline urine which lowers tubular reab-sorption of salicylic acid and increases excretion,which could be the basis of the findings by Nanra andKincaid-Smith [70], viz, that addition of bicarbonateto a regime of plain aspirin (500 mg/kg/day) reducesthe incidence of renal papillary necrosis by one half.

Bluemle and Goldberg [1061 found that dosesbetween 170-300 mg/kg of aspirin or 300 mg/kg ofphenacetin or paracetamol given separately or to-gether (170mg of aspirin plus 300mg ofphenacetin perkg of body wt), acutely with or without hydration,were handled differently by the kidney. At the time ofpeak concentrations in the blood, cortical and medul-lary assays were performed. No medullary gradientsfor salicylate was found, whether the animals weredehydrated or not, but free and conjugated paraceta-mol concentration rose sharply in the inner medulladuring dehydration. The gradient was more than 10:1

for conjugated paracetamol and 19:1 for free parace-tamol. Salicylate did not affect the paracetamol gra-dient. The pH dependency of salicylate clearanceand distribution was not determined.

Duggin and Mudge [107, 108] have confirmed thework of Bluemle and have overcome objectionsregarding the contribution made by tubular fluid inthe medulla; they also confirm that paracetamol isconcentrated in the medulla and that dehydrationincreases the gradient. Paracetamol is sparinglybound to protein(c.f., aspirin) and is therefore filteredby the glomerulus and reabsorbed by the tubule bysimple diffusion; this is true for its conjugates whichat low concentrates are actively secreted and at highconcentrations reabsorbed (dog). Phenacetin is notconcentrated in the medulla.

The data with regard to the medullary gradient foraspirin is conflicting. Bluemle and Goldberg [1061found no gradient in the dog, but Gault [1091 demon-strated a gradient for aspirin in rabbits and guineapigs (1:7, rabbit; 1:3, guinea pig) and a modest gra-dient for paracetamol (1:2.7, rabbit) and for phenace-tin (1:3.6, rabbit). The radioactivity of the labelledaspirin in the kidney was 20 times that of the liver,spleen, pancreas, muscle, and brain.

Quintanilla and Kessler [110] gave sodium salicy-late to dogs (3 mmoles/kg) and also described a med-ullary gradient for salicylate of a factor of 1:5 to 3(see their table 7). This, however, could be artefac-tual, as the medullary concentration was exactlyequal to the plasma concentration, and the "gra-dient" was due to very low concentrations in thecortex. These workers also found that saucy lateadministration induced a prompt diuresis with excesssecretion of sodium, potassium, and chloride, andoccasionally glucose, which is the reverse of theeffects found in man.

Little data is available on the kidney/blood ratiofor salicylate in patients; however, in two patientsdying of salicylate poisoning, the kidney blood ratioswere 8.0/15.0 and 82.4/0.6; such a variation makesany conclusion impossible [106, 1101.

Whitehouse [Ill] has shown that when guineapigs were pretreated with aspirin at 200 mg/kg/day,blood concentrations of a single dose of paracetamol,(150 mg/kg) were higher; and more significantly,excretion of paracetamol by the kidney was inhibit-ed, although sulphation and glucuronidation wereunaffected. Comparable results have been found inman [1121. Whitehouse believes that aspirin com-petes preferentially for the common anionic secreto-ry mechanism in the kidney tubule.

Finally, Carro-Ciampi [113] has shown that a 50%reduction in posttreatment plasma concentrations ofphenacetin is obtained with the first ten days of

Analgesic nephropathy: Drug mechanisms 19

repeated phenacetin treatment at a dose of 0.4 g/kglday in albino rats; in the guinea pig, posttreatmentphenacetin concentrations fell to only approximately20%. Paraphenetidine levels increased as treatmentproceeded in guinea pigs, but it fell in rats.

In man it has been shown that: 1) The absorption,excretion, and metabolism of phenacetin and aspirindoes not vary with the tubular cell response [114].2) The recovery of phenacetin and its metabolitesis essentially the same whether patients have analge-sic nephritis or not, but the renal excretion of con-jugated paracetamol is reduced when nephropathyis present, and this correllates with reductions inthe glomerular filtration rate. The metabolism ofphenacetin, in contrast to aspirin, is not pH depen-dent [115]. 3) Steady state salicylate concentrationsare decreased when aspirin is given chronically (20days) in healthy volunteers. The average plasma con-centration at day 21 was approximately half that ofday 7 [116]. 4) Met-hemoglobin formation [1171 inabusers of analgesic mixtures containing phenacetinis significantly greater than in normal subjects orpatients with impaired renal function. The relevanceof met-hemoglobin formations to renal disease per seis disputed, however. 5) There is evidence that thelimiting pathway on the formation and excretion ofmetabolites of aspirin [118], i.e., salicyluric acid andsalicylphenolic glucuronide, does not increase pro-portionately with dose (from 65 mg/kg to 100 mg/kgdaily).

Indomethacin plasma concentrations in man [119]following a standard dose of 100 mg are significantlyincreased when buffered aspirin is given before orconcurrently.

Van Ginneken [120] has performed the most com-prehensive study of pharmacokinetics of commonlyused analgesics in healthy volunteers. Ibuprofen,aclofenac, fiufenamic acid, mefenamic acid, phena-zone and its isopropyl and methylamino derivatives,paracetamol, phenacetin, acetylsalicylic acid, andsodium salicylate were given as single or multiple(ibuprofen, aclofenac, acetylsalicylic acid) doses for7 to 14 days. Chemical methods were used for esti-mation of plasma and urine concentrations; in addi-tion to plasma and urine kinetics, an estimation ofprotein binding and red cell uptake was made.

Ibuprofen and aclofenac had small volumes ofdistribution, high protein binding, but short half-lives(due to small volume of distribution). Ibuprofen wasexcreted almost unchanged in the urine (c.f., ado-fenac, 50%) and secreted almost unchanged by thetubule in contrast to aclofenac. No accuminulation ofeither occurred, but both showed marked fluctua-tions in plasma concentrations following repeatdosing.

The fenamates showed irregular plasma curves,which could not be fitted to any simple kinetic mod-el. The renal excretion was negligible and the biologi-cal availability of the tested formulations was low;although flufenamic acid was better absorbed thanmefanamic acid.

The pyrazoles showed large inter-individual varia-tions and large volumes of distribution, and thegroup showed large differences in clearance,although as a whole, renal clearance played only aminor role. 4-Aminophenazone is eliminated quitedifferently from the others in the group; 40% is elimi-nated partly by renal excretion (refer to the thesis[120] for detailed discussion).

Phenacetin produced very low plasma concentra-tions due to high hepatic clearance, in contrast toparacetamol, and the pharmaceutical formulation ofphenacetin produced marked differences in theabsorption.

Salicylates showed the same volume of distribu-tion as did ibuprofen and aclofenac, the plasmakinetics were non-linear, and at higher doses clear-ance decreased; both tubular secretion and reabsorp-tion occurred which confirm earlier findings. The pHdependency of excretion was once again demonstrat-ed in chronic studies: doubling the daily dose ofsodium salicylate from 1.5 to 3 g led to a four to five-fold increase in average plasma plateau concentra-tion, which fits well within the model of capacity-limited elimination.

The reader should refer directly to this excellentmonograph which shows clearly the virtue of repeatplasma kinetics after acute and repeated dosing[120].

Extrapolation of findings from animals to man:Species differences. The survival of any animal spe-cies reflects the adaption, in both form and function,to its habitat. Animal experiments have been per-formed mainly in rats, but also in guinea pigs, rab-bits, cats, dogs, baboons, monkey, and pigs. Allhave important structural and functional differencesin comparison to man. This is true even of animalswhose kidney architecture and physiology mustclosely resemble that of man (probably the pig andthe monkey).

The rat kidney has only a single papilla, the rela-tive length of which is greater than man's, so that thenumber of long nephrons is higher than man [95,121]. In man, the number of short looped nephrons is86% (c.f., pig, 97%), and the papilla is better sup-plied with blood, compared to the rat. The supply inthe rat is postglomerular only, but the edges of thehuman papilla are supplied from aglomerula vessels,first appearing in intrauterine life and becoming moreimportant as man ages. The central core of the papil-

20 Shelley

la, too, has an aglomerula supply, so the loss ofglomeruli with aging does not imperil the blood sup-ply to such an extent. The medullary flow in man isless well studied, but it is likely that the main meta-bolic pathway in the medulla is the anaerobic utiliza-tion of glucose [1171. There is a high sympatheticinnervation to the rat efferent arteriole, to the upperparts of the vasa rectae and to the juxtamedullaryefferent arteriole which (in the rat) contain contrac-tile cells; the sympathetic nerve supply to the humanis not well studied, however. In summary, the kidneyof the rat has major architectural differences to thatof man; the latter is, if anything, better able to with-stand anoxia.

The strain used is a further factor. Boyd [122] hassummarized the etiologic factors affecting phenace-tin toxicity in the rat, and seasonal variation in sus-ceptibility has also been noted repeatedly by Kin-caid-Smith. Male Manor Farm rats have been shownto be much more susceptible to an indomethacinderivative than are Charles River CD rats [94]. Wis-tar rats have a higher incidence of natural renalmalformations which makes long-term studies diffi-cult to interpret. The Gunn rat [123, 124] has beenparticularly susceptible to the development of papil-lary necrosis from aspirin in doses which did notaffect Wistar rats at all (200 to 500 mg/kg aspirin; 500and 190 mg/kg/day APC). The reasons for this havebeen discussed by Axelsen and Burry [125]. Sex-related susceptibility has been shown in the case ofsudoxicam [127] (where the plasma concentration ismuch higher in females), and age-related susceptibili-ty has been shown with a derivative of indomethacin[94, 102, 122]. In many studies, only female Wistarrats have been used, e.g., Nanra et al [96], Saker etal [134], and in other studies the sex is not stated.Findings in different animal species have beenreviewed by others [94, 101, 126, 127].

As the pig has a multi-papillary kidney and theability to concentrate urine in a manner comparableto man, it is of interest that the pharmacokinetics ofsalicylates in pig and man are comparable. Firststudies of analgesic nephropathy are appearing now[95]. Wiseman and Reinhart [127] summarized thesituation: "renal papillary necrosis has been pro-duced mainly in rats, occasionally in rabbits anddogs, and rarely in monkeys," although Lechat,Levillian, and Dechezlepaetre [52] reported that gla-phenine causes tubular lesions in baboons. Thelesions produced by phenylbutazone and indometha-cm in cat and dog have already been referred to.There seems no good reason why these speciesshould be used in preference.Proposed mechanisms for the genesis of the renal lesion

1) Direct toxic effect by concentration in inedul-

lary tissue. The consensus of opinion is that the firstlesion in experimental analgesic nephropathy occursin the long loop of Henle and vasa recta, areas wherehigher concentrations of (some) compounds occur.Maximum concentrating ability may be lost in theabsence of detectable structural damage, and there isevidence from animal experiments that this occurs inexperimental analgesic nephropathy.

The data regarding cortico/medullary gradients forcommonly used analgesics has been summarizedearlier, and there is direct evidence that dehydrationenhances the severity of the experimental lesionfrom high doses of aspirin and APC, and the medul-lary gradients where they exist, although this doesnot hold for studies when lower doses were used.

Indirect support to the direct toxic effect of analge-sics has been provided by the elegant experiments ofHardy [123, 124] and Brown and Hardy [101].Although removal of the papilla in the rabbit [124]produced severe damage per se, this was probablydue to the technique employed and the ensuingischemia. This can be avoided, and the partiallypapillectomized kidney (rat) functions as an intactorgan. When a fenamic acid derivative is given tosuch rats (known from previous experiments to pro-duce renal papillary necrosis consistently), renalpapillary necrosis occurs only in the intact kidney(i.e., not in the contralateral partially papillectom-ized kidney); the cortical lesion also occurred in theintact kidney, suggesting the papilla to be the prima-ry site of damage. The interesting possibility arises; ifthe site of concentration of analgesics is removed,does this protect the kidney against further directtoxic damage, while maintaining the ability to con-centrate urine at least under conditions of mild fluiddeprivation?

In man the shedding of the normal papilla wasthought to carry a poor prognosis, but Kingsley et al[11] stated that there is "a little evidence that it hasany permanent effect on renal function."

2) The role of anoxia. The oxygen tension in thehuman kidney is known to fall from 100 mm Hg inthe cortex to about 20 mm Hg in the medulla [128],Bernanke and Epstein [129] have shown that in the(hydropenic) dog, the medulla receives only one-tenth to one-twentieth of the cortical supply. Themedulla, therefore, is clearly vulnerable to ischemiain both species.

Anoxia may result from the following changes inthe vascular tree: a) reduction in flow by vasocon-striction or mesangial thickening, b) platelet aggre-gates, c) occlusion of blood vessels by interstitialhyperplasia, d) changes in the oxygen affinity ofhemoglobin, e) changes in viscosity.

Nanra et al [96, 130] have reviewed data on vascu-

Analgesic nephropa thy: Drug mechanisms 21

lar damage, and they have demonstrated that APC(900 mg/kg/day), aspirin (500 mg/kg/day), paraceta-mol (3,000 mg/kg/day), caffeine (150 mg/kg/day),phenylbutazone (10 mg/kg/day), mefanamic acid (100mg/kg/day), all lowered medullary flow, mefanamicacid being the most active in this respect. Theischemia produced correlated with impaired medul-lary function, but not with the histological lesions.This work is so important that it should be repeatedwith control groups in each subset, as a number ofmaneuvers (dehydration, shock, diuresis) wereemployed which should be stratified.

Endothelial necrosis and vascular obliterationhave also been reported in animals [97, 1311, but notin man. Brown et al [101] reported minor vesselchanges in rats, but Abrahams [1321 reported a spe-cific microangiopathy in patients. Nanra and Kin-caid-Smith [96] also reported platelet aggregates inthe vasa recta, which is parallel to the findings ofAbrahams, Rubenstein, and Levin [131] but not tothose of Arnold et a! [102] and Ham and Tangue[133]. Saker and Kincaid-Smith [134] observedadherent platelets, but no thrombi. In view of theknown effects of the nonsteroidal antiinflammatoryagents on platelet aggregation and adhesion, it wouldseem very unlikely that they would induce aggrega-tion or thrombosis.

Blood flow may be reduced by occlusion of vesselsby surrounding tissues. The data here is conflicting:Gault, Blennerhasset, and Muehreke [67] confirmedthe earlier work of Clausen by showing an increase ininterstitial connective tissue with an increase in col-lagen about the walls of the vasa recta and tubule.These changes were observed early in the disease.Molland observed quite the opposite—damage anddisappearance of interstitial cells [135].

Salicylate administration reduces erythrocyte 2,3-diphosphoglyceride in a dose-related manner, whichcorrelates well with the plasma concentration. Suchreduction increases the affinity of hemoglobin tooxygen, thereby depriving tissues; and Kravath,Abel, and Colli [136] have quoted data suggestingthat salicylates increase venous oxygen saturation by10 to 20% in rat and man.

3) Metabolic effects. The main metabolic pathwayin the inner renal medulla consists chiefly of theanaerobic utilization of glucose. Aspirin has beenshown to inhibit the hexose monophosphate shunt inmedullary tissue [117] (unlike paracetamol). In theisolated rat tubule [137], salicylate administrationreduced intracellular ATP [137], resulting in areduced rate of gluconeogenesis; however, in thewhole medulla of the dog, Quintanella and Kessler[110] found medullary ATP to be unaffected andcomments that this is to be expected, as the metabo-

lism there is primarily glycolytic. Adenosine triphos-phate (ATP) fell and adenosine monophosphate(AMP) rose in the cortex, indicating that aspirin,in this region, does uncouple oxidative phos-phorylation.

Davidson, Daffist, and Shippey [138, 139], usingdog renal medullary slices in aerobic or anaerobicconditions, showed that only when salicylic acid andparacetamol were given together was substratemetabolism altered, and that salicylic acid inhibitedamino acid incorporation into protein. Paracetamolhas little effect alone, but it enhances that of saucy-late. This matter is discussed further in other articles[6—8, 71, 140, 141].

4) Immunological. Little data is available. Murrayand von Stowasser [142] postulated, on the basis ofglomerular deposits and positive staining of glomer-uli incubated with fluorescent antiserum, that there isan immunologic basis, that the amino group con-tained in most analgesics may be responsible, andthat these findings parallel to some extent the earlierviews of Harrow. In human studies, however, Gaultet a! [67] found no IgC or f3lC globulin in the cortexor medulla of six patients with early analgesicnephropathy.

5) Prostaglandins. The renal papilla is the mainsource of prostaglandins in the kidney. As "aspirin-like" drugs inhibit prostaglandin biosynthesis, couldthis property be relevant to the genesis of analgesicnephropathy? (See [143—148].)

The degree of inhibition of prostaglandin synthe-tase by the nonsteroidal antiinflammatory drugsdepends very much on the substrate employed [144,145]. The PG2 series are derived (by definition) fromarachidonic acid. In the rabbit medulla only a 2 to3% conversion has been observed; in the cortex,none. Despite this low level of metabolism in therabbit, however, high levels of A2, E2, and F2 in themedulla of the rabbit have been measured and posi-tively identified by mass spectrometry and gas chro-matography [1491. In the rat, there is histochemicalevidence for the existence of prostaglandin hydro-genase in the tissue surrounding the ascending loopof Henle, and there is circumstantial evidence ofprostaglandin synthesis in the dog also.

The rank order of inhibition of synthetase in theguinea pig lung by the nonsteroidal antiinflammatorydrugs is as follows: meclofenamic acid > niflumicacid > mefenamic acid > flufenamic acid > naprox-en > phenylbutazone > aspirin, or ibuprofen [144,150, 151]. Fjalland eta! [152] would place indometh-acm before flufenamic acid, and Ku, Wasuary, andCash [153] would place diclofenac as the most potentof all. Aspirin and indomethacin inhibit synthetase inthe rabbit medulla, producing inhibition of E2 and

22 Shelley

F2, also. Paracetamol appears to inhibit brain syn-thetase only.

a) Modulation of the effect of reninlangiotensin.Angiotensin and noradrenalin increase renal perfu-sion pressure and the basal output of prostaglandin-like materials in the rabbit and rat; when the synthe-sis of the latter is blocked by indomethacin, theeffects of the autonomic neurotransmitters are aug-mented or reduced depending on the experimentalsituation. When indomethacin is given intraarterially(1 to 2pg/ml) the magnitude of duration of theresponse to the pressor agents is reduced [1541. Thepressor effects of noradrenalin were enhanced whenPGE2 was infused into the kidney in amounts whichby themselves had no pressor activity, and Gagnon,Gaulhier, and Regoli [1551 showed that when thetissues are pretreated with indomethacin (2g/m1)the increases in perfusion pressure induced by angio-tensin and noradrenaline were significantly greater.

Nerve stimulation, too, releases prostaglandins,and indomethacin can block the effects of nervestimulation and produce a rise in systemic bloodpressure in every case and total renal blood flow insome experiments [1491.

Vasoconstriction per se releases prostaglandins.Two receptors for angiotensin are proposed. Onemodifies prostaglandin secretion, the other is con-cerned solely with vasoconstriction [155].

In dogs, indomethacin reduces renal blood flowfrom 15 to 45% (as does aspirin) and causes redistri-bution from the inner to the outer cortex. Berg dis-cusses reasons why the effects are more pronouncedin human patients with renal failure than in normalsubjects [73, 74].

b) Direct vascular effects. Although PGF2c, ismildly pressor, it has negligible effects on renal vas-cular resistance. PGE2 on the other hand producesvasoconstriction in the isolated perfused kidney andreduces renal blood flow (in contrast to the earlierfindings of Lee, quoted by Arnold [1031) Simpleinhibition of this compound, therefore, would notexplain the reduction in medullary blood flowobserved by Nanra and Kincaid-Smith [1301. Infu-sions of arachidonic acid produces hypotension inguinea pigs. Aspirin will inhibit this effect [156].Pretreating animals with salicylic acid before admin-istering aspirin, however, produces a normalresponse to arachidonic acid, i.e., salicylic acid pre-vents inhibition by aspirin.

c) Natriuresis. Short-term administrations of aspi-rin and indomethacin produce sodium and waterretention. Donker, Arisz, and Brentjens [75] believe,in the case of indomethacin, that this is a conse-quence of inhibition of prostaglandin synthesis,which in turn leads to an activation of the renin-

angiotensin system. High doses of angiotensin, how-ever, markedly decrease rat renal tubular electrolyteresorption, thereby inducing natriuresis. This effectcan be enhanced by indomethacin [157], i.e., urineflow and sodium excretion increase in thisexperiment.

Bartelheimer and Senft [158] also failed to showwater and sodium retention during administration ofindomethacin (8 mg/kg) in rats, although phenylbuta-zone and oxyphenbutazone did do so. It is clear,then, that indomethacin produces varying responseswith regard to renal blood flow and glomerular filtra-tion rate in different species and depending on theexperimental situation.

The natriuretic effect of pro staglandins couldresult from an effect on oxidative metabolism. Lee[159] has shown that prostaglandins inhibit interme-diary metabolism which, by inhibiting sodium andpotassium ATPase, leads to a diminution of highenergy intermediates available for sodium transportand, thus, produces natriuresis.

The reader is referred to the papers of Berg 1173,74], Davis and Horton [149], and Gagnon, Gaulhier,and Regoli [155] for a fuller discussion. It is clear,however, that the vascular changes induced by pros-taglandins are complex and their inhibition by non-steroidal antiinflammatory agents are comparably so.The data in man shows acute sodium and waterretention following administration of an analgesic,e.g. aspirin, which could be explained by an inhibi-tion of the natriuretic effects of pro staglandins; how-ever, during chronic administration of analgesics,inappropriate sodium loss is present.

Summary

The situation in the experimental field is unre-solved; too many factors require clarification beforethe critical experiment can be conducted to settle thematter once and for all. However, as there is nowplentiful evidence to convince any reasonable physi-cian that commonly available analgesics, whenabused, carry a significant health risk, one may rea-sonably ask whether any further experimental evi-dence is needed?

The object of this review is in no sense divisive,i.e., by pointing out discrepancies in the availabledata to thereby cloud the issue rather than resolvethem. The problem of abuse lies properly in the fieldof public health education, and the first step to thiswould surely be an appropriate warning on the pack-aging of all commonly used analgesics.

For future research, however, government healthauthorities should be guided in their preclinical test-ing requirements for mild antiinflammatory analge-

Analgesic nephropathy: Drug mechanisms 23

sics, and enough is now known to draw up guidelinesfor good laboratory practice in this field.

Acknowledgments

I acknowledge the help and advice of Dr. L. F.Prescott; the views expressed, however, and anyerrors therein are the writer's responsibility. Mrs. C.Boyett typed and assisted in revising the manuscript.

Reprint requests to Dr. J. H. Shelley, C. H. Boehringer Sohn,Southern Industrial Estate, Bracknell, Berkshire RGJ2 4YS, Unit-ed Kingdom.

References

I. GSELL 0: From phenacetin nephritis to analgesic nephropa-thy 1953—1974. Schweiz Rundschau Med (Praxis) 63:1299,1974

2. RINGOIR S: Aspects cliniques, epidemiologiques et experi-mentaux de Ia nephropathie chronique avec abusd'analgesiques. Therapie 29:483—505, 1974

3. PRESCOTT LF: Phenacetin nephropathy. Br Med J 4:493,1970

4. PRESCOTT LF: Aspirin and renal papillary necrosis. Br MedJ 2:468, 1971

5. MURRAY RM, LAWSON DH, LINTON AL: Analgesic nephro-pathy: clinical syndrome and prognosis. Br MedJ 1:479—482,1971

6. MURRAY T, GOLDBERG M: Chronic interstitial nephritis:etiologic factors. Ann Intern Med 82:453, 1975

7. MURRAY T, GOLDBERG M, Analgesic Abuse and renal dis-ease. Ann Rev Med 26:537—550, 1975

8. NANRA RS, KINCAID-SMITH P: Chronic effect of analgesicson the kidney. Progr Biochem Pharmacol 7:285—323, 1972

9. NANRA RS, KINCAID-SMITH P: Renal papillary necrosis inrheumatoid arthritis. Med J Aust 1:194—197, 1975

10. ARGER PH, BLUTH E, MURRAY T: Analgesic abuse nephro-pathy. Urology 7:123, 1976

11. KINGSLEY DPE, GOLDBERG B, ABRAHAMS C, MEYERS AJ,FURMAN K, COHEN I: Analgesic Nephropathy. Br Med J4:656—659, 1972

12. Editorial: Analgesic nephropathy or phenacetin poisoning.Br MedJ 2:588, 1974

13. Editorial: Analgesic nephropathy and aspirin. NZ Med i79:1073. 1974

14. Editorial: Analgesics and the kidney. Br Mcdi 3:123, 197315. Editorial: Analgesic nephropathy: Some unsolved problems.

Med JAust 1:185, Feb. 15, 197516. SEWART JH, Mc CARTHY SW, STOREY BG, ROBERTS BA,

GALLERY E, MAHONY iF: Diseases causing end-stage renalfailure in New South Wales. Br Mcdi 1:440—443, 1975

17. KRAMER P: Analgesic nephropathy. Med K/in 70:889—895,1975

18. SHELLEY JH: Phenacetin through the looking glass. C/inPharmacol Therap 8:427—471, 1967

19. SMITH JD, RABKIN R, STABLES D: Analgesic renal papillarynecrosis. S Afr Mcdi 49:1819—1822, 1975

20. GOLDBERGER LE, TALNER LB: Analgesic abuse syndrome.Urology 5:728—732, 1975

21. GOLDBERG M, Mc CANNON D: Phenacetin and papillarynecrosis. Br Mcdi 1:348. 1971

22. WIGLEY RAD: The New Zealand Experience. Aust NZ iMed 6 (suppl. l):37—39, 1976

23. Committee on Safety of Medicines, Dept. of Health & Social

Security, United Kingdom, Personal Communication31.1.77

24. PRESCOTT LF: Anti-inflammatory analgesics and drugs usedin the treatment of rheumatoid arthritis and gout, in Mey/er'sSide Effects of Drugs, edited by DUKES MNG. Amsterdam,Excerpta Medica, 1972— 1975, vol. 8, pp. 207—240

25. PRESCOTT LF: Antipyretic analgesics, in Meyler's SideEffects of Drugs, edited by DUKES MNG, Amsterdam,Excerpta Medica, 1972— 1975, vol. 8, pp. 154—206

26. ABEL JA: Analgesic nephropathy: A review of the literature1967—70, Clin Pharmacol Ther 12:583, 1971

27. DUBACII UC, ROSNER B, MULLER A: Relation betweenregular intake of phenacetin-containing analgesics and labo-ratory evidence for uro-renal disorders in a working femalepopulation of Switzerland. Lancet 1:539, 1975

28. GLOOR F: Current concepts on morphology and pathogen-esis of analgesic nephropathy. Schweiz Med Wochenschr104:785, 1974

29. HOFFLER D: Renal insufficiency due to phenacetin. DtschArztebl 21:1447, May, 1976

30. KRIKLER DM: Paracetamol and the kidney. Br Mcdi 2:6 15,1967

31. Mc DONALD FD, LAZARUS GS, CAMPBELL WL: Phenylbu-tazone anuria. Southern Mcdi 60: 1318, 1967

32. BLOCH-MICHEL H, GORINs A, MEYEROVITCH A: The renalincidents of phenylbutazone. La Presse Medico/c 74:2671,1966

33. MORALES A, STEYN J: Papifiary necrosis following phenyl-butazone ingestion. Arch Surg 103:420—421, 1971

34. TOL0T F, PROST G, GUIGNOT B: Multiple signs of intoler-ance to an anti-inflammatory drug association: Role of phen-ylbutazone. Scm Hop Paris 51:2955—2960, 1975

35. BRAUN J: Abnormalities of urinary sediment and renal failurefollowing sulfinpyrazone therapy. Arch Intern Med 136:1060,1976

36. MARSH FP, ALMEYDA JR, LEVY IS: Non-thrombocytopenicpurpura and acute glomerulonephritis after indomethacin.Ann Rheum Dis 30:501, 1971

37. PIETRAK C: Intensive treatment in acute renal failure afterindomethacin poisoning. Rheumato/ogia 12:433—435, 1974

38. PETER E: Double-blind comparison of naproxen and indo-methacin on the after-midnight pain of patients with ankylos-ing spondylitis. Arzneim Forsch 25:324, 1975

39. Ross FA: A double-blind comparison between diclofenacsodium and ibuprofen in Osteoarthritis. i mt Med Res 3:267,1975

40. THOMPSON M, CRAFT AW, AKYOL MS, PORTER R: Ibupro-fen in the treatment of rheumatic diseases: Long-term experi-ence with observations on lack of adverse effects. Curr MedRes Opinion 3:594, 1975

41. MOLONY J, PIGGOT PV, QUILL JO: A double-blind trial of anew anti-inflammatory drug in the, management of osteoar-thritis of the hip joint. Jlrish Med Assoc 64:605, 1971

42. DOURY P, PATTINS 5: Study of a new anti-inflammatory drug(Ketoprofen) in rheumatology (206 cases). Rheumatologie2:185, 1973

43. SAMARA A, MARQUES JF: Ketoprofen (19.583 RP) in thetreatment of rheumatoid arthritis. Rev Bras Cl/n Therap 4:89,1975

44. PHILLIPS BM, LEELING JL, K0wAL5KI TL: Does aspirinplay a role in analgesic nephropathy. Aust NZJ Med 6 (suppl.1):48—53, 1976

45. BROWN JH, HALL J, BIUNDO JJ: Results of a one-year trialof tolmetin in patients with rheumatoid arthritis. J Cl/n Phar-macol 15:455—463, 1975

24 Shelley

46. BILLINGS RA, BURRY HC, EMSLIE FS, KERR GD: Vasculi-tis with aiclofenac therapy. Br Med J 4:263, 1974

47. AYLWARD M, PARKER RJ, HOLLY F: Long-term study ofindomethacin and alciofenac in treatment of rheumatoidarthritis. BrMedJ 2:7, 1975

48. BARNES CG, GOODMAN HV, EADE AW: A double-blindcomparison of naproxen with indomethacin in osteoarthritisJ C/in Pharmacol 15:347—354, 1975

49. Bmzus G, JOSENHANS G: Results of an open therapeutictrial with naproxen in patients with degenerative joint diseas-es. Scand J Rheumatology (Suppl.) 2:80—88, 1973

50. LARCAN A, CALAMAL M, LAMBERT H, GRIJNINGER C: Poi-soning by glafenine with acute renal failure. Ann Med Nancy11:507—508, 1972

51. LECHAT P: The consumption of analgesic antipyretics inFrance. Therapie 29:553—560, 1974

52. LECHAT P, LEVILLIAN R, DECHEZLEPAETRE S: Demonstra-

tion of acute nephrotoxicity of glafenine in the rat. Therapie29:567—573, 1974

53. DEQUIEDT P, VANHILLE P, DEVALOER B: Acute renal insuf-ficiency due to glafenine. Lille Med 20:708—714, 1975

54. GAULTIER M, BISMUTH C, MONEL-MAROGER M: Acuteinterstitial nephritis due to glafenine poisoning. Nouv PresseMed 1:3125—3128, 1972

55. GAULTIER M, BISMUTH C, MONEL-MAROGER M: Acuteinterstitial nephritis in glafenine poisoning. Therapie 29:579—585, 1974

56. GAULTIER M, BISMUTH C, MONEL-MAROGER M: Acute

tubulo-interstitial nephropathy during poisoning by glafenine.Europ J Toxicol 5:310, 1972

57. CHIVRAC D, MARTI R, FOUNIER A: Immuno-allergic haemo-lysis complicated by acute renal insufficiency after ingestionof glafenine. Nouv Presse Med 3:2578, 1974

58. CHEVET D, RAMEE MP, GARRE M: Acute poisoning withglafenine; Acute tubulo-interstitial nephropathy. Europ JToxicol 7:306—309, 1974

59. STORK J: Glafenine and nephrotoxicity. Ned T Geneesk

120:338, 197660. DUPLAY H, MATTEL M, BARILLON D: Acute tubular and

interstitial nephritis in glafenine poisoning. Therapie 29:593—

597, 197461. KATONA G: Niflumic acid, a new nonsteroidal anti-inflam-

matory agent: Clinical experiences in degenerative diseasesof joints. Rev Med Hosp Gen (Mexico) 34:251—259, 1971

62. BERATO J: Analysis of 100 cases of longterm treatment withnifiumic acid. Gaz Med Fr 79 (suppl.5): 10, 1972

63. CAMP0DONIcO A, DIAZ J, Dii' 0, HULT E, SIMONDI JJ:Acute renal failure caused by drugs. Rev Med Rosario 58:16,1970

64. RIFFAT G: Bucloxic acid in rheumatic conditions. Gaz MedFr 82:1705—1708, 1975

65. Roux 0: Clinical trial of a new anti-inflammatory agent inO.R.L. J Franc Ow Rhino Laiyng 24:545, 1975

66. VILLAUMEY J, DI MENZA C, ROTTERDAMM M: The treat-ment of pain in osteoarthntis by short courses of niflumicacid. Sem Hop Paris (Ther), 50:355—361, 1974

67. GAULT MH, BLENNERHASSET J, MUEHRCKE RC: Analgesic

nephropathy: A clinicopathologic study using electronmicroscopy. Am J Med 151:740—756, 1971

68. MURRAY RM: Persistent analgesic abuse in analgesicnephropathy. J Psychosomatic Res 16:57—62, 1972

69. MURRAY RM, SMITH R: Coexistent analgesic nephropathyand bromism. Lancet 1:73—74, 1972

70. NANRA RS, KINCAID-SMITH P: Salicylates and papillarynecrosis. Br Med J 2:45, 1971

71. NANRA RS: Nephrotoxicity of aspirin. Bull Post grad CommMed Univ Sydney 29:221, 1973

72. NANRA RS: Pathology, aetiology and pathogenesis of analge-sic nephropathy. Aust NZJ Med 6 (suppl. l):33—37, 1976

73. BERG KJ: Acute effects of acetylsalicylic acid in patientswith chronic renal insufficiency. Europ J C/in Pharmacol11:111—116, 1977

74. BERG KJ: Acute effects of acetylsalicylic acid on renalfunction in normal man. Europ J C/in Pharmacol 11:117—123, 1977

75. DONKER AJM, ARISZ L, BRENTIENS JR: The effect of indo-methacin on kidney function. Kidney Jut 7:362, 1975

76. HUVAR A, SITAR J: Phenylbutazone, ketophenylbutazone.Vnitrni Lekarstvi 2: 147—153, 1971

77. HARBST H, KLINKE JD: Pharmacologic influence of p-hydroxyphenylbutazone on the isotope renogram. Nuci Med(Stuttg.) 8:191—194, 1970

78. BEELEY L, KENDALL MJ: Effect of aspirin on renal clear-ance of 1251-Diatrizoate. Br Med J 1:707—708, 1971

79. ROBERT M, FILLASTRE JP, BEYER H, MALANDAIN H:Effect of intravenous infusion of acetylsalicylic acid on renalfunction. Br MedJ 2:466-467, 1972

80. KENDAL MJ, NUTTER S, HAWKINS CF: Xylose Test: Effectof aspirin and indomethacin. Br Med J 1:533—536, 1971

81. BURRY HC: Reduced glomerular function in rheumatoidarthritis. Ann Rheum Dis 3 1:65, 1972

82. BURRY HC, DIEPPE PA, BRESNIHAN FB, BROWN C: Salicy-lates and renal function in rheumatoid arthritis. Br Med J1:613—615, 1976

83. LEATHWOOD PD, PLUMMER DT: The excretion of lactatedehydrogenase in human urine after the ingestion of aspirin.BiochemJ 114:197—202, 1969

84. DIEI'PE PA, DOYLL DV, BURRY HC, TUCKER SM: Renaldisease in rheumatoid arthritis. Br MedJ 1:611—612, 1976

85. SALANT D, WILJOEN R, MILNE J, RABKIN R: Tubular func-tion in analgesic nephropathy. Kidney mt 8:121, 1975

86. CAUGHEY DE, and 13 others: Aspirin and the kidney: NewZealand Rheumatism Association Study. Br Med J 1:593—596, 1974

87. EDWARDS KDG, SCHAPEL GJ, JEREMY R, STEELE TW:Possible nephrotoxicity of aspirin in rheumatoid arthritis.MedJAust 1:492, 1972

88. BURRY HC, DIEPPE PA: Apparent reduction of endogenouscreatinine clearance by salicylate treatment. Br Med J 2:16—17, 1976

89. ELLIOT HC, MURDAUGH HV: Effects of acetylsalicylic acidon excretion of endogenous metabolites by man. Proc SocExp Biol Med (N. Y.), 109:33—335, 1962

90. COVE-SMITH JR. KNAPP MS: Sodium handling in analgesicnephropathy. Lancet 2:70, 1973

91. EDWARDS OM, EDWARDS P. HUSKISSON EC, TAYLOR RJ:Paracetamol and renal damage. Br Med J 2:87—89, 1971

92. Coronary Drug Project Research Group: Aspirin in coronaryheart disease. J Chron Dis 29:625, 1976

93. BENGTSSON U:Phenacetin-containing analgesics and chronicrenal disease. Lancet 1:1195, 1975

94. BOKELMAN DL, BAGDON WJ, MATTIS PA: Strain-depen-dent renal toxicity of a nonsteroid anti-inflammatory agent.Toxicol Appi Pharmacol 19:111—124, 1971

95. MCIVERMA, HOBBS J: The failure of high doses of aspirin toproduce renal lesions in pigs. MedJ Aus 1:197—199, 1975

96. NANRA RS, KINCAID-SMITH P: Medullary ischaemia inexperimental analgesic nephropathy: The pathogenesis ofrenal papillary necrosis, in Phenacetin Abusus 1973, editedby HASCHEK H, Vienna, Facta Publication, 1973, pp. 45—88

Analgesic nephropathy: Drug mechanisms 25

97. NANRA RS, KINCAID-SMITIl P: Papillary necrosis in ratscaused by aspirin and aspirin-containing mixtures. Br Med J3:559—561, 1970

98. ARNOLD L, COLLINS C, STARMER GA: Renal and gastriclesions after phenylbutazone and indomethacin in the rat.Pathology 6:303—313, 1974

99. SEsSA A, ALLARIA PM, CONTE F: Ultrastructural changes ofthe glomeruli of the rat induced by indomethacin. Nephron10:238—245, 1973

100. CARLISLE CH, PRESCOTT RH, PENNY CW: Toxic effects ofphenylbutazone on the cat. Br Vet J 124:560, 1968

101. BROWN DM, HARDY TL, COLLINS C, STARMER GA: Short-term study of the effect of phenacetin, phenazone and amido-pyrine on the rat kidney. Br J Pharmacol Chemother 32:17—24, 1968

102. ARNOLD L, HARDY TL, COLLINS C, STARMER GA: Theshort-term effects of analgesics on the kidney, with specialreference to acetylsalicylic acid. Bull Post-Grad Comm MedUniv Sydney 29:214—219, 1973

103. ARNOLD L, HARDY TL, COLLINS C, STARMER GA: Theshort-term effects of acetylsalicylic acid in the rat kidney:Further observations. Nephron 10:355—375, 1973

104. ADAMS SS, BOUGH RG, CLIFFE EE, LE55EL R, MILLSRFN: Absorption, distribution and toxicity of ibuprofen.Toxicol AppI Pharmacol 15:310—330, 1969

105. LEELING JL, PHILIPPS BM, KOWALSKI TC: Influence ofaspirin formulation and dose on concentration of total salicy-late in rat kidney. J Pharmacol Sci 65:447, 1976

106. BLEUMLE LW, GOLDBERG M: Renal accumulation of salicy-late and phenacetin: Possible mechanisms in the nephropathyof analgesic abuse. J C/in Invest 47:2507—2514, 1968

107. DUGGIN GG, M'JDGF GH: Analgesic nephropathy: Renaldistribution of acetoaminophen and its conjugates. J PharmacolExpPathol 109:1, 1976

108. DUGGIN GG: Renal tubular transport of paracetamol and itsconjugates in the dog. Br J Pharmacol 54:359—366, 1975

109. GAULT MH: Demonstration of an intra-renal gradient for 14Caspirin in rabbits and guinea pigs. C/in Res 19:808, 1971

110. QUINTANELLA A, KESSLER RH: Direct effect of salicylateson renal function in the dog. J Clin Invest 52, 3143, 1973

111. WHITEHOU5E LW, PAUL CJ, THOMAS BH: Effect of aspirinon fate of '4C-acetaminophen in guinea pigs. J Pharmacol Sci64:819, 1975

112. AMSEL LP, DAVISON C: Simultaneous metabolism of aspirinand acetaminophen in man. J Pharmacol Sci 61:1474—1475,1972

113. CIAMPI G: Tolerance to repeated phenacetin treatment inalbino rats and guinea pigs. Toxicol AppI Pharmacol 22:641—648, 1972

114. PRESCOTT LF, SANSUR M, LEVIN W: The comparativemetabolism of phenacetin and N-acetyl-p-aminophenol inman, with particular reference to effects oii the kidney. C/inPharmacol Therap 9:605—6 14, 1968

115. PRESCOTT LF: The metabolism of phenacetin in patients withrenal disease. Clin Pharmacol Therap 10:383—394, 1969

116. MULLER FO, HUNDT HKL, KOCK AC: Decreased stead-state salicylic acid plasma levels associated with chronicaspirin ingestion. Curr Med Res Opin 3:417, 1975

117. GAULT MH, SHADHIDI NT: Methemoglobin formation inanalgesic nephropathy. C/in Pharm Therap 15:521—527, 1974

118. GIBSON T, ZAPHIROPOULOS G, GROVE J: Kinetics of saucy-late metabolism. Br J Clin Pharmacol 2:223—238, 1975

119. GARNHAM JC, HICKS DC, RAYMOND K: Some observationson the influence of buffered aspirin on the absorption ofindomëthacin in man (abstr.). Br J C/in Pharmacol 1: 178p,1974

120. VAN GINNEKEN CAM: Pharmacokinetics of antipyretic andanti-inflammatory analgesics. Ph.D Thesis, Published by Stu-dentenpers, Nijmegen, Holland, 1976

121. FOURMAN J, MOFFAT DB: The Blood Vessels of the Kidney.Oxford, Blackwell Scientific Publications, 1971

122. BOYD EM: A review of phenacetin tocicity in animals.Applied Therapeutics 12:9, 1970

123. HARDY TL: N-phenyl anthranilic acid: An agent for inducingand studying renal papillary necrosis in the rat. Br J ExpPathol 51:348, 1970

124. HARDY TL: Partial renal papillectomy and the study of druginduced renal papillary necrosis in the rat. Br J Exp Pathol51:591, 1970

125. AXELSEN RA, BURRY AF: Papillary necrosis in the Gunn rat:Rapid induction by analgesics. Patho/ogy 4:225—229, 1972

126. ROSNER J: Nephropathies experimentales aux analgesiques.Therapie 29:507—546, 1974

127. WISEMAN EH, REINERT H: Anti-inflammatory drugs andrenal papillary necrosis. Agents and Actions 5:322-325, 1975

128. LEONHARDT KO, LANDES RR: Oxygen tension of the urineand renal structures. N Engl J Med 269:115—121, 1963

129. BERNANKE D, EPSTEIN FH: Metabolism of the renal medul-la. Am J Physiol 208:541, 1965

130. NANRA RS, CHIRAWONG P. KINCAID-SMITH P: Medullary

ischaemia in experimental analgesic nephropathy: The patho-genesis of renal papillary necrosis. Aust NZ J Med 3:580—586, 1973

131. ABRAHAMS C, RUBENSTEIN AH, LEVIN NW: Phenacetininduced papillary damage in experimental animals. Nature200:695, 1963

132. ABRAHAMS C: Cause of analgesic-induced renal papillarynecrosis. Lancet 2:346, 1976

133. HAM KN, TANGUE JD: Some experimental observationsrelating to the pathogenesis of renal papillary necrosis. AustAnn Med 18: 199—208, 1969

134. SAKER BM, KINCAID-SMITH P: Papillary necrosis in experi-mental analgesic nephropathy. Br MedJ 1:161—162, 1969

135. MOLLAND E: Aspirin damage in the rat's kidney in the intactanimal and after unilateral nephrectomy. VI mt CongrNephrol, Florence, abstr. 774, 1975

136. KRAVATH RE, ABEL G, COLLI A: Salicylate poisoning:Effect on 2,3-diphosphoglycerate levels in the rat. BiochemPharmacol 21:2656—2658, 1972

137. DAWSON AG: Effects of acetylsalicylate on gluconeogenesisin isolated rat kidney tubules. Biochem Pharmacol 24:1407,1975

138. DAVIDSON W, DAFFIST L, SHIPPEY W: Effect of aspirin andphenacetin metabolies on protein synthesis in dog renalmedulla. C/in Res 21:683, 1973

139. DAVIDSON W, DAFFIST L, SHIPPEY W: In vitro effects ofaspirin and phenacetin metabolies on the metabolism of dogrenal medulla. C/in Res 21:227, 1973

140. BARRACLOUGH MA: Effect of vasopressin on the reabsorp-tion of phenacetin and its metabolites from the tubular fluid inman. C/in Sci 43:709—713, 1972

141. BURRY HC: The kidney and arthritis. Proc Roy Soc Med66:897, 1973

142. MURRAY G, VON STOWASSER V: Glomerular lesions inexperimental renal papillary necrosis. BrJExp Pathol 57:23—29, 1976

143. COLLIER HOJ: Prostaglandins and aspirin. Nature 232:17,1971

144. FLOWER Ri: Drugs which inhibit prostaglandin biosynthesis.Pharmacol Rev 26:1974

145. FLOWER RJ: Aspirin, prostaglandins, endoperoxides andplatelets. Nature 253:88, 1975

26 Shelley

146. FERREIRA SH, MONCADA S, VANE JR: New aspects of thcmode of action of nonsteroid anti-inflammatory drugs. AnnRev Pharmacol 14:57—73, 1974

147. McGIFF JC, CROW5HAW K, TERRAGNO A, L0NIGR0 AJ:Renal prostaglandins: Their biosynthesis, release, effects,and fate, in Renal Pharmacology, edited by FISHER JW,London, Butterworths, 1971

148. SMITH MJH, FORD HIJTCHISON AW, ELLIOT PN: Prosta-glandins and the anti-inflammatory activities of aspirin andsodium salicylate. J Pharm Pharmac 27:473—478, 1975

149. DAvIs HA, HORTON EW: Output of prostaglandins from therabbit kidney, its increase on renal nerve stimulation and itsinhibition by indomethacin. Br J Pharmacal 46:658—675,1972

150. DEREADT R: Inhibition of prostaglandin biosynthesis by non-narcotic analgesic drugs. Arch ml Pharmacadyn 224:30—42,

1976151. DEBY C, DESCAMP5 M, BINON F: Correlation between anti-

inflammatory properties and inhibition of prostaglandin bio-synthesis in vitro. Biachem Pharmacal 24:1089, 1975

152. FJALLAND B, WILJOEN R, MILNE J, RUBKIN R: Inhibitionby non-steroidal anti-inflammatory agents of the release ofrabbit aorta contracting substance and prostaglandins fromchopped guinea pig lungs. J Pharm Pharmacal 26:448—451,1974

153. Ku EC, WA5UARY JM, CASH WD: Diclofenac sodium (OP45840, voltaren), a potent inhibitor of prostaglandin synthe-tase. Biochem Pharmacal 24:641—643, 1975

154. ARMsTRONG iF, BI,ACICWELL GJ, FLOWER Ri: Possiblecontribution of prostaglandins to genetic hypertension inrats: Identification of a biochemical lesion. Br J Pharmacol35:244, 1975

155. GAGNON Di, OALJLFIIERR, REGOLI D: Release of prosta-glandins from the rabbit perfused kidney: Effects of vasocon-strictors. Br J Pharmac 50:553—558, 1974

156. LEFORT J, VARGAFTIG BB: Salicylate acid prevents inhibi-tion by aspirin of arachidonic acid-induced hypotension,bronchoconstriction and thrombocytopenia. Br J Pharmacol60:292p, 1977

157. OREINER KG, KEMPER R, OSWALD H, SCHMLTZ Hi: Poten-

tiation of angiotensin Il-induced natriuresis by indomethacinin the rat. Br J Phar,nacol 59:435-436p, 1977

158. BARTELHEIMER HK, SENFT G: Localisation of the tubulareffect of some antirheumatic agents. Arz Forsch 5:567, 1968

159. LEE iB: The anti-hypertensive and natriuretic endocrinefunctions of the kidney: Vascular and metabolic mechanismsof the renal prostaglandins, in Prostaglandins in Cellularbiology, edited by RAMWELL PW and PHARRI55 BG, NewYork, London, Plenum Press, 1972, pp. 399—449